philip lelyveld The world of entertainment technology


OpenAI’s GPT-3 Wrote This Short Film—Even the Twist at the End

Now there’s another feat to add to GPT-3’s list: it wrote a screenplay.

It’s short, and weird, and honestly not that good. But… it’s also not all that bad, especially given that it was written by a machine.

The three-and-half-minute short film shows a man knocking on a woman’s door and sharing a story about an accident he was in. It’s hard to tell where the storyline is going, but surprises viewers with what could be considered a twist ending.

The students that created the film used a tool derived from GPT-3 called Shortly Read to write the screenplay. They wrote the first several lines then let the AI take over.

According to media futurist and algorithmic filmmaker Alexis Kirke, we should get used to the idea of computers having a hand in our creative endeavors, especially when it comes to writing movie scripts. “A huge amount of experience has been codified by writers, producers, directors, script editors, and so forth,” he told Digital Trends“Want to reduce the number of adverbs and adjectives in your script? There’s an algorithm for that. Want to ensure your characters’ dialog all sound[s] different from each other? There’s an algorithm for that. Want to generate alternative, less clichéd, rewrites of a page that keep its general meaning? There’s an algorithm for that.”

See the full story here:


Ethics vs. compliance in AI

But when it comes to responsible technology use, Heather Roff, senior research analyst, the Johns Hopkins University Applied Physics Laboratory, warned that viewing it as compliance could be detrimental. 

"Ethics is not compliance and if we think about ethics as compliance then we are failing as moral agents," Roff said during the panel. 

"If you think about ethics as compliance or compliance officers or inspectors general then you are actually getting the bare minimum of what ethics is. Ethics is about…how to think about doing things responsibility. What am I thinking about when I build it" and not "is the compliance officer watching."

Roff also said basic research and testing were integral to the ethics conversation. 

"Basic research is just not funded to the levels it needs to be funded," Roff said. "We need more basic research and funding for things like testing and evaluation...because testing supports the principles."

See the full story here:


Carnegie Mellon U Big Brain Joe Marks Joins Weta Digital As Chief Technology Officer

Marks has over 35 years in as a technology executive with background in graphics, computer science and research and development. Marks will oversee the company’s technology initiatives that span visual effects as well as the newly announced Weta Animated division, and will report to Akkaraju, who said the company was grateful to “have attracted technology leadership of Joe Marks’s caliber to Weta Digital. A pivotal part of the future of VFX and animation rests in artificial intelligence, machine learning and cloud computing. Joe’s deep experience in these areas makes for a perfect fit at Weta as we continue to push the boundaries of our industry.”

Marks joins Weta Digital from Carnegie Mellon University, where he served as Executive Director of the Center for Machine Learning & Health, working at the forefront of innovation in digital healthcare. Marks is best-known for his prior roles at Mitsubishi Electric Research Labs (MERL) and Disney Research where he orchestrated and led the premier R&D organization in the media and entertainment industry with labs in Pittsburgh, Zurich, Boston, San Francisco and Los Angeles. Marks’s research interests span CG, video processing, media distribution, robotics, HCI, mobile computing, computer vision, AI, sensors, embedded systems and behavioral economics.

“While working with The Walt Disney Company and the ACM SIGGRAPH community, I admired Weta for many years,” Marks said. “It’s an absolute thrill to be joining the company as it embarks on a new adventure in animated storytelling, while continuing their unparalleled tradition of excellence in VFX for live-action movies. 

See the full story here:


Saudi Arabia Sets out Strategy to Become a World Leader in Artificial Intelligence By 2030

The implementation of the National Strategy for Data & AI will follow a multi-phase approach, which will focus on addressing the immediate national requirements for Saudi Arabia until 2025, particularly the Vision 2030 strategic development goals. From 2025 until 2030, Saudi Arabia will focus on developing specialisms in specific areas of AI, to become one of the leading AI and data-driven economies by 2030.

View source version on

See the full story here:


AI can be used in all retirement plans if focused properly – WPS panelists

"We need to really think about what are the safe spaces to onboard AI. Recognize that our industry is not very good at failing, and yet there is no innovation without failure. … That's hard in our world where we have prudent person rules and fiduciary duties. We're not designed for that," Mr. Monk said.

Plans should find safe areas to experiment with AI, he said. "The first place to apply AI is not to try to generate alpha. It would be, let's think about our operational footprint, see if we can unleash AI to take a bunch of unstructured data and make it structured … to do better passive investing." Once there is comfort around the role AI can take in an investment process, then look into how it can be integrated "in attempts to beat the market," Mr. Monk said.

See the full story here:


New Photoshop tool could help fight fake images

 On Tuesday, Adobe announced the release of a "beta" version of an attribution tool for Photoshop, with the hope it will encourage people to trust that the images they see are truthful (or at least make more informed judgments).The feature, which is optional and will be available initially just to select users, lets photo editors append images with detailed information known as metadata that, in essence, travels around with it online. This information will go far beyond the basic details that can currently be added to pictures and may include who created the image and where, a thumbnail of the original image, and data about how it has been altered — as well as whether AI tools were used to change the picture. This data will be secure and it will be clear if it has been tampered with, Adobe (ADBE)said.

The release is part of the company's Content Authenticity Initiative to fight against dis- and misinformation, which Adobe launched a year ago with Twitter and The New York Times. Initially, the content attribution tool will be for publishing still images to Behance, an Adobe-owned social network for sharing creative work. Over time, the company hopes this kind of authenticating information will be added to different types of content, and be shared widely on social media platforms and through media companies.

Adobe's effort will be limited by the fact that those editing images need to use it voluntarily. And the company won't be the first to attempt to popularize a method for making media trustworthy. However, it may have a better chance at success than others due to its reach. 

Yet, as Farid pointed out, Adobe's content-attribution tool can't confirm the veracity of an image before it is edited in Photoshop. To address this problem, photo- and video-verification startup TruePic, which is part of the Content Authenticity Initiative and for which Farid is a paid adviser, recently announced it partnered with chip-maker Qualcomm to create a way to securely snap pictures via a smartphone's the built-in camera app.

See the full story here:


It’s Time to Try Virtual Reality. Here’s Where to Start.

PhilNote: this story is a basic sampler.

The rapidly expanding VR universe, which is accessible from an app store within the headset, will please beginners of all tastes and ages, from an experienced PC gamer to your 8-year-old niece. Here are some tips on where to begin.

See the full story here:


Deconstructing Maxine, Nvidia’s AI-powered video-conferencing technology

Maxine, a platform that leverages artificial intelligence to improve the quality and experience of video-conferencing applications in real-time.

Maxine used deep learning for resolution improvement, background noise reduction, video compression, face alignment, and real-time translation and transcription.

In this post, which marks the first installation of our “deconstructing artificial intelligence” series, we will take a look at how some of these features work and how they tie-in with AI research done at Nvidia. We’ll also explore the pending issues and the possible business model for Nvidia’s AI-powered video-conferencing platform.

See the full story here:



Construction on the $1.2 billion UMUSIC Broadwater hotel complex in Biloxi, Mississippi – the first of Universal Music Group’s recently announced “music-based experiential hotel properties” – will commence next year and will be completed in 2023.

UMUSIC Broadwater will also have a 12,000-capacity music venue.

UMUSIC Broadwater in Biloxi, Mississippi is first of three locations for the hotel series, with other locations to be announced for Atlanta, Georgia and Orlando, Florida. There are also plans for additional locations across the US and globally.

See the full story here:


Bookful, World’s Largest Augmented Reality Library App, Welcomes Mattel’s Barbie® and Thomas & Friends™

It’s been more than half a century since the world was introduced to Barbie, the world's number 1 fashion doll. The iconic brand is so much more than a doll, with over 200 careers and counting, Barbie continues to inspire the limitless potential in every girl, and with new titles including Barbie: You Can Be a Doctor and You Can Be a Designer Bookful will empower young readers to bring to life Barbie’s ethos “You Can Be Anything”.

Thomas & Friends was created 75 years ago by Reverend Wilbert Awdry when he sought to entertain his young son, Christopher, while he was recovering from the measles. Today, Thomas is more than just an engine with hundreds of TV episodes, books, toys and even a theme park.

The Barbie and Thomas & Friends titles are joining Bookful’s library of top publishers and brands including Penguin Random House, Charlesbridge and DK.

Bookful by Inception brings children’s books to life with AR and 3D animations providing productive screen time while cultivating a love of reading and learning. With hundreds of books and educational games, Bookful is the world’s largest 3D/AR library.

See the full story here: